There is epidemiological evidence of a close association between major alteration in glucose metabolism and breast cancer risk. In three prospective studies there was a doubling of breast cancer risk for women who had a diagnosis of diabetes at baseline. In a recent prospective cohort study, we found a strong association of fasting glucose with risk of breast cancer. Overweight and obese women with breast cancer have poorer survival compared with thinner women. A recent study conducted at three Toronto hospitals observed that fasting insulin was associated with breast cancer outcomes in 512 women with early stages of the disease. The aim of the proposed study is to evaluate the long term prognostic significance of fasting glucose in primary breast cancer. The study will be conducted on 20,432 women admitted to the Istituto Nazionale per lo Studio e la Clara dei Tumori (Italian National Cancer Institute) in Milan, Lombardy Region, Italy between 1991 and 2002 for surgical treatment of a primary breast cancer. Women included in this study will be residents of the Lombardy Region (Northern Italy), who received complete resection of the neoplastic lesion (lumpectomy with margin clear of invasive cancer or mastectomy). We will derive information on pre-treatment serum fasting glucose levels from a computerized file of the Clinical Laboratory of the Italian National Cancer Institute in Milan. Personnel of the Lombardy Cancer Registry (LCR) will carry out the complete follow-up of the study participants for breast cancer recurrence and death. In addition to serum glucose, we will be able to collect individual information on total and HDL-eholesterot, triglycerides, and serum uric acid from the same Clinical Laboratory file. These factors are, together with serum glucose, constituents of the Syndrome X, a condition defined by a cluster of metabolic factors and indicated to be a marker of insulin resistance and impaired glucose metabolism. Thus, we also propose to study the potential relation of individual pre-treatment baseline exposure to Syndrome X and breast cancer outcomes. Body weight and age, menopausal status, presence of diabetes at admission, traditional prognostic factors, such as tumor size, nodal stage, estrogen-progesterone receptor status and treatment related variables will also be available in the clinical chart and reported in the file together with the Clinical Laboratory data and they will be included in the analysis as potential confounders, or effect modifiers.